Project Summary Osteoarthritis (OA) is the leading cause of disability worldwide. The inability of non-invasive techniques to quantify disease progression has limited understanding of the pathogenesis of OA. While numerous magnetic resonance imaging (MRI) methods have been proposed for imaging OA, sensitivity to bone metabolism has been limited. We propose to develop advanced three-dimensional PET-MRI methods for bone and soft tissue metabolism to study the response of the tissues in the joint to changes in knee load. This work will lead to a new understanding of OA pathogenesis by revealing relationships between changes in cartilage and bone metabolism over time. This project aims to develop PET-MRI methods to sensitively track changes of OA in response to biomechanical loading. Our specific aims are to (1) Develop accurate, reproducible and dose-optimized kinetic models of dynamic 18F-NaF PET-MRI for quantitative bilateral whole joint imaging using deep learning and advanced MR coil technology, (2) Study the relationship between resting state bone metabolism and biomechanics using PET- MRI and (3) Perform a longitudinal study to assess the response of our new imaging methods to changes in joint biomechanics from gait retraining. The innovation of this work lies in the development of novel imaging techniques that simultaneously offer quantitative measures of tissue physiology in cartilage and bone using PET-MRI. The significance of this work is that we will be able to sensitively and quantitatively track changes in bone metabolism and soft tissue microstructure due to changes in biomechanical loading in the knee joint over time. This will provide new and more sensitive imaging tools to assess the responses of the joint to biomechanical interventions to treat OA such as gait retraining, bracing, or high tibial osteotomy.